Interviewer Ali: Describe to me the phenomenon of cache breakdown, and talk about your solution?

Foreword

Cache (memory or Memcached or Redis.....) is widely used in Internet projects. This blog will discuss the topic of cache breakdown, covering the phenomenon of cache breakdown, solutions, and codeAbstract way to handle cache breakdown.

What is cache breakdown?

​

The above code is a typical way of writing: when querying, first get it from the Redis cluster, if not, then query it from the DB and set it to the Redis cluster.Note that in actual development, we generally store in the cache, and the data structure stored is JSON.(The serialization method provided by JDK is slightly less efficient than JSON serialization; and JDK serialization is very strict, and the increase or decrease of fields may cause deserialization failure, and JSON has better compatibility in this regard) Suppose query from DBIt takes 2S, so obviously the requests that come during this period of time will all go through the DB query under the above code, which is equivalent to the direct penetration of the cache. This phenomenon is called "cache breakdown"!

Analysis of ideas to avoid cache breakdown

Add synchronized?

​

If synchronized is added to the method, all query requests have to be queued. Originally, our intention was to let concurrent queries go to the cache.That is, the granularity of synchronized is too large now.

Reduce the granularity of synchronized?

​

In the above code, when there is data in the cache, the request for querying the cache does not have to be queued, which reduces the granularity of synchronization.However, the problem of cache breakdown is still not resolved.Although multiple DB query requests are queued, even if one DB query request is completed and set in the cache, other DB query requests will continue to query DB!

synchronized+double check mechanism

​

Through the synchronized + double check mechanism: In the synchronized block, continue to judge and check to ensure that it does not exist, and then check the DB.

Code abstraction

No, in fact, the code we deal with the cache is templated except for the specific query DB logic.Now let's abstract it!

An interface for querying DB:

​

Since querying the specific DB is determined by the business, then expose this interface for the business to implement it.

A template:

​

Don't Spring have a lot of Template classes?We can also abstract the code through this idea, let the outside world decide the specific business implementation, and write the template steps.(somewhat similar to the concept of AOP)

Improved code:

It can be seen from this that we don't care where the cached data is loaded from, but to the specific user, and the user no longer has to pay attention to the problem of cache breakdown when using it, because we give abstraction.